Plant seeds as protein sources of food or feed. Evaluation based on

Cecil H. Van Etten, William F. Kwolek, John E. Peters, Arthur S. Barclay. J. Agric. Food Chem. , 1967, 15 (6), pp 1077–1089. DOI: 10.1021/jf60154a01...
0 downloads 0 Views 1MB Size
Plant Seeds as Protein Sources for Food or Feed. Evaluation Based on Amino Acid Composition of 379 Species C. H. V a n E t t e n , ' W. F. Kwolek,' J. E. Peters,' and A. S. Barclay? Amino acid analyses are reported for the first time o n seeds from 165 plant species. F o r evaluation, the data were combined with results reported earlier on 214 other species. Based o n the F A 0 provisional pattern for selected nutritionally essential amino acids, the seed proteins were generally adequate in leucine, phenylalanine, threonine, and valine. The mean methionine content for 35 species of Compositae, and the mean lysine content for 92 species of Cruciferae, and for 70 species of Leguminosae, were

he amino acid compositions of seed from 214 angiospermous species have been reported (Miller et d., 1962a,b; VanEtten et d.,1961b, 1963a). The data supplied basic information on the distribution of the nitrogenous constituents in a variety of plant seeds. This paper presents results on the amino acid composition of seeds from 165 additional species representing 136 genera and 47 plant families. The combined data from 379 species support generalizations made from the present authors' previous work. In addition, plant seeds are evaluated as a source of food protein. based on their amino acid composition. Appraisals are made by comparison with the essential amino acid provisional patterns recommened by the Food and Agriculture Organization of the United Nations (FAO), (1957; WHO 1965). Such an evaluation is desirable as a part of the search for additional sources of food. Predictions of future food needs based on current rate of world population increase and food production emphasize the seriousness of this problem (Altschul, 1965; F A O , 1964; Hamilton. 1965). From studies such as that of F A 0 (1964) it appears the most urgent need is to increase production of protein, especially of good nutritional quality. In the past the most practical source of food including protein has been cereals and other harvested seeds. Advantages are that food in seeds is in a concentrated, easily preserved form, and that seed crops can often be grown close to where they are needed. Such sources of food are being increased through development of superior varieties. better agronomic practices, and improved harvesting, storage, and processing methods. Perhaps more food can be produced from lesser known domestic plants or by domestication of wild ones. New food crops may be developed that are adapted to areas now considered marginal for agriculture. Also. as the compositions of less familiar plants become known, we can be

T

1 Northern Regional Research Laboratory, Agricultural Research Service, U .S. Department of Agriculture, Peoria, Ill. Crops Research Division, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, Md.

above the mean for all species. Percentages of methionine and isoleucine in seeds from most plant families were below F A 0 requirements. Seeds from Gramineae (including the common cereal grains) were low in lysine. Seed proteins from a number of species have a better pattern of essential amino acids than many crop seed sources. Many seed meals contained toxic or deleterious substances which must be inactivated or removed before the meals can be considered for food or feed.

more confident of finding new sources of raw material for industry. Information presented here is germane to such attempts. MATERIALS A N D METHODS

Seed preparation, hydrolysis of seed proteins, and ion exchange chromatographic analyses of amino acids were carried out as described by VanEtten et d.(1961b). Selections of the species were made from a n analytical compilation of more than 1650 samples studied for their content of crude protein, oil, and other components (Earle and Jones, 1962; Jones and Earle, 1966). Amino acid analyses were determined by the automated method of Spackman, Stein, and Moore (1958) with a model MS Beckman Spinco instrument. The 150-cm. column was operated at 30" and 50" C . in order to separate hydroxyproline from aspartic acid. Amino acid compositions cover species from plant families as follows: Leguminosae, 24; Cruciferae, 19; Compositae, 18; Labiatae, 10; Boraginaceae and Euphorbiaceae, five each; Rutaceae, four; Amaryllidaceae. Apocynaceae, Polemoniaceae, and Umbelliferae, three each; and the remaining 68 species from 53 diferent plant families. Of the angiosperms, seven are monocots and 154 dicots. F o r the first time. four species of the gymnosperms are reported. AMINO ACID COMPOSITION A N D VARIABILITY

Mean Composition and Variation from the Mean. Crude protein and oil content of the seed, seed plus pericarp, or seed minus seed coat for the 379 species calculated on the dry basis were: crude protein (nitrogen X 6.25), mean 27.9 %, extremes 5.6 to 71 .0 %, standard deviation 10.2; oil, mean 26.9%, extremes 0.8 to 66.0%, standard deviation 15.2. The means for each amino acid (Table I) for the 379 species are similar to those already reported for 200 species by VanEtten et a/. (1963a). The greatest changes in variation, as measured by relative standard deviations, are in arginine, lysine. phenylalanine. and tyrosine (see last column of Table I). F o r all 379 species, lysine, methionine, arginine, glycine, phenylalanine, tyrosine, glutamic VOL. 15, NO. 6, N0V.-DEC. 1967

1077

Table I. Amino Acid Lysine Methionine Arginine Glycine Histidine Isoleucine Leucine Phenylalanine Tyrosine Threonine Valine Alanine Aspartic acid Glutamic acid Proline Serine

Summary of Amino Acid Compositions of Seed from 379 Speciesa Grams per 16 Grams Nitrogen Relative Standard Deviation Mean Extremes Std. Dev. 379 species 200b species Change

4.39 7.5-1.3 1.19 1.56 3,5~-0,5 0,469 8.58 20.1-3.1 2.57 4 85 12.6-2.6 1 .oo 2.27 4.3--1.2 0.396 3.60 5.8-1.9 0.543 6.05 13.7-3 2 0.947 3.87 10.1-2.0 0.792 2.88 5.3-1.6 0.601 3.31 5.0-1.6 0.562 4.52 6.7-2.3 0.698 3.96 8.8-1.5 0,709 8.41 14.5-4.2 1.51 16.76 33.1-8.6 3.36 4.33 11 . 3 - 1 . 1 1.27 4.12 6.7-2,3 0.645 il Statistical terminology recommended by Anal. Chent. (1961). From amino acid composition of seed meals from 200 plant species (VanEtten et a/., 1963a).

27.1 30.1 29.9 20.6 17.4 15. I 15.7

24.2 29.8 26 8 19.4 15.0 15.3 16.7 16.3 16 8 15.9 15 6 18.7 17.6 20.8 27.5 16.5

20 5 20 9 17.0 15.4 17.9 17.9 20.0 29.3 15.7

2.9 0 3 3. I 1.2 2.4 0 2

1 .o

4.2 4.1 1.1 0.2

0.8 0 3

0 8 1 8 0 8

~ . . _ _ _

acid, and proline varied the most (relative standard deviation range of 20.0 to 30.1). The remaining amino acids have a relative standard deviation range of 15.1 to 17.9. In this summation, hydroxyproline was not included, as it has been found only in such tissues as seed coat and pericarp (VanEtten et al., 1961a). Of the 165 species reported in Table 11, the 10 species analyzed as seed kernels without seed coat or pericarp contained n o hydroxyproline. This analysis is additional evidence that storage and embryo protein d o not contain hydroxyproline. The frequency distributions for lysine, isoleucine, and valine (Figure 1) appear symmetrical about the mean. When scattered outliers of the extreme greater than the mean are not considered, the distribution also appears symmetrical for methionine, leucine, and phenylalanine. A frequency distribution for the remaining amino acids often showed a wide scatter of outliers at the extreme greater than the mean. A possible explanation is that seed from a n occasional species contains a less familiar or unidentified amino acid or nitrogenous base which elutes with the known amino acid. However, the two high leucine values are from the few species of Gramineae reported in this study. By microbiological assay Taira (1962a,b; 1963) reports leucine content in this range for members of the Gramineae, subfamily Panicoideae. Deyoe and Shellenberger (1965) give high leucine values for grain sorghums. Amino Acid Composition in Relation to Plant Groups. The mean analytical results shown in Table I11 were calculated from data o n 379 species. The arithmetic mean and the number of samples involved are given for all families sampled five or more times. At the bottom of the table is listed the standard deviation per observation calculated from variability between samples from the same family. Sources of variation involved in the standard deviation include variation among genera, species, and determinations, as well as environmental effects. Comparing means for different families is complicated by unequal numbers of samples and unequal numbers of genera and species in families. Even more important, the number of 1078 J. AGR. FOOD CHEM.

I-I ::PLA lysine C . 2.8 R . 4.4 s . 6.8 W . 2.7 P 4.2

140

Methionine c . 2.0 R . 2.1 S . 1.5 W . 1.6 P . 2.2 AP . 1 . 8

-

'

Isoleucine C . 4.0 R . 4.5 S . 5.1 W . 3.6 __ P . 4.2

leucine C . 13.9 R . 8.8 s . 7.7 W - 6.9 P . 4.8

-

:100

-,G

I I )

n

0

1

3

5

Tyrosine c 4.0 R . 2.4 s . 3.9 W . 3.1 P . 2.8 '

-

7

1

2

3

Phenylalanine C . 5.2 R . 5.7 S . 5.0 W . 4.7 P . 2.8

-

2

3

4

5

Threonine c . 3.6 R . 3.8 s . 4.3 W . 2.9 P . 2.8

Valine

C . 5.0 R . 5.6 s 5.4 W . 4.4 P . 4.2 '

140

,1_ 2

3

4

5

Amino Acid Content, g./16 g. N

Figure 1. Frequency distribution of nutritionally essential amino acids in plant seeds from 379 species compared with the estimated requirement for man (FAO, 1957) and with the amino acid content of corn (Xlertz et a/.,1965), rice (Cagampang et a/.,1966) soybeans (Rackis et ([I., 1961), and wheat (Waggle et a/., 1967)

Requirements for methionine and for phenylalanine are based on protein containing 2 or more cystine and 2.8 or more tyrosine. respectively; with less clstine and tyrosine, more methionine and phenylalanine would be required: C = corn; R = rice; S = soybeans; W = wheat: P = F A 0 (1957) provisional pattern: A P = WHO (1965) adjusted provisional pattern

01

m w

m. m

N. e>

c

s s

2

r - x

-r-

b.

oc 30

d

7 7 ' 9 m

0

d

m

.

.

2 I

c

x

c

-

c,

P>

2

2

r-

N

c

x

d

u

e.

N

.

s

e, N

-

e.

h

x

.

"

.

r

-

N

-. x

7

m m

r-

s

a

moc

r-

x

icic3.

?

d

d

d

G

e i

a

I n b G

e a.

i. s

G.

-.

N

N

h

N

N

-

*

x. -

w,

u u l n .

.

G

.

9

c-

VOL. 15, NO. 6, N0V.-DEC. 1967 1079

N

dl

m N

e

N

f)

E

c" au!J na1os1

n c

d

1080 J. A G R . FOOD CHEM.

7 c,r-, 0 "; c, c '9 3, 0 0 c c,7 c,'9 0 p c 0 c,'? ccccccecc3cccccc30cccccc

VOL. 15. NO, 6, N0V.-DEC. 1967 1081

1082

J. AGR. FOOD CHEM.

c N m

c N c d i i h &r.i r - w r- r- 3 r - r - w r -

-

r-.

-.

h3

nrjc r-mr-

d d

w, m, is, d.

3 m

X

-r-

c r-

CO

00

3

m

- m N ?

N. X. - m ,

.

h

3"0,hZ

c --w-, N

id:i m v;

i d P P

&

?m ev,-cc,? ? d.m . Nee m. m. m-x

w m

60

"i 7 c

-'"ye, c

c c c --oca

c

.

- N -

-NN

. . .

xxr-

C t \ d d O I*, e,m m m d 6 N X I A

d

Pi

m

N vi

r - m c 2 -

r:aaaa 3 . m - - -

L"Nr.i

"icxr-? dvrv'dd r -. e .m i s .m .

- - -

2x-n-

- 3 x

. . .

-

31 G e, PI h l

VOL. 15, NO. 6, N0V.-DEC. 1967

1083

x

P-

m

a

-

3:

w

00

.-

-

m

m

m

9 m

N m

W

-

c

0

N

z

?

'ti

3\

vi

d

r-

3

m

d:

m

?

vi

r,

m

h

00

N

N

-it

n

m

-

vi

N

cc,

m

I-

N

vi

m

m m

P-

VI

'9

N

N

m

m

Ti

b

P-

?

N

P-

I-

-

?

0

N

N

N

N

w

%

m

CCI

N a

c

? 3

-

c. b

2

2

3

.Y

53

e

:

1084

J. AGR. FOOD CHEM

2 u

c. C i

n

h

N

c.r.

X

m

n

s

E

E

In N

X cc

cc

r-

2 :.

N r-

. ss

,

m

r-

d

.

r-

-

C

c.1

m

bmm

s

F, N

.

x s m

SI

X

"p"'3 ? .b: . x - - -

d

C G a

I

VOL. 15, NO. 6. N0V.-DEC. 1967 1085

samples in many instances includes less than 5 % of the genera in the family. As a rough measure in making family comparisons two least significant differences (LSD) are given a t the end of Table 111 for use when each of two means being compared is based o n either five or nine observations. If the difference exceeds the appropriate LSD, the means differ significantly with approximately one chance in 20 that the difference actually results from sampling variability. Examples of relationships between plant family and composition seen in Table 111 are: the low lysine content of seed from Gramineae and Rosaceae; the high lysine in Cruciferae and Leguminosae; the low methionine content of Leguminosae and Rosaceae; the high arginine and glycine in Cucurbitaceae and the low amount of these amino acids found in Gramineae. Isoleucine is low in Malvaceae and Rosaceae; leucine and phenylalanine are high in Gramineae; threonine is high in Cruciferae; valine is high in Euphorbiaceae; alanine and glutamic acid are high in Gramineae; proline is high in Gramineae and Cruciferae. Based o n analyses of seeds of 54 species of Gramineae, Taira (1962a,b, 1963) found that the subfamilies Pooideae, Eragrostoideae, and Panicoideae differ from each other in their amino acid patterns and also from the subfamilies Pharoideae and Arundinoideae which have similar amino acid patterns. Seeds from the Panicoideae, which include corn and sorghum, were much higher in leucine and alanine than seed from the other subfamilies. The means for the amino acids from the 54 species were: lysine 3.40, methionine 1.76, arginine 4.40, glycine 4.00, histidine 2.10, isoleucine 4.39, leucine 9.34, phenylalanine 4.84, tyrosine 2.12, threonine 3.79, valine 5.60, alanine 6.90, aspartic acid 6.69, glutamic acid 22.1, proline 11.80, and serine 5.02. These means agree rather well with those for only five species from the Gramineae reported in Table 111, except for tyrosine and proline. The relative standard deviations of each of the amino acids in seed from 14 species of Lesquerella (Miller ef ai., 1962b) were less than those for each of the amino acids from 41 species representing 29 genera of the Cruciferae (Miller et a/., 1962a) except for leucine, alanine, and serine. This variation is consistent with expected greater uniformity within genera than within the family. The mean for each amino acid within the genus Lesquei-ellri was about the same as the mean for the 41 species from the 29 genera of the Cruciferae, except for lysine. Different varieties of Brassica canipestris and of B. ncipus grown in different places, as indicated by analysis of four accessions from each species, showed few, if any, differences in amino acid composition among the accessions or between the species.

such as albizziin (Gmelin, 1959), and probably amino acids that have not been characterized. Other sources of nitrogen are nonprotein amino acids stable to acid hydrolysis such as canavanine, which is tentatively identified from its elution position, and a,P-dianiinopropionic acid, which elutes with histidine under the conditions of analysis, Others appear as unidentified elution peaks. Some may go undetected because they elute in the same position as one of the protein amino acids, and thus are erroneously calculated as part of it. Currently, of the average 10 new amino acids characterized per year according to Meister (1965), a high percentage have been isolated from plants. In the Cruciferae and related plants, thioglucosides containing nitrogen are found (Kjaer, 1960). On acid hydrolysis the nitrogen in these compounds probably form ammonia (VanEtten et ai., 1963a). Recent reports show that the organic aglycon of the thioglucosides has a biogenetic source in common with the amino acids (Benn, 1965; Chisholm and Wetter, 1964). Canavanine in the Leguminosae. Canavanine contents of seed from the Leguminosae are given in Table IV. None of these have been previously reported as containing the compound, except Meliiotus alba. Of 70 Leguminosae analyzed in this survey, 23 contained canavanine ranging from 0.7 to 18.7 grams per 16 grams of nitrogen. The compound is present also in other Leguminosae (Bell and Tirimanna, 1965; Birdsong et ai., 1960). Unidentified and Less Familiar Amino Acids. 4-Hydroxy-L-pipecolic acid has been isolated from Acacia seeds (Leguminosae) by Virtanen and Gmelin (1 959) and from Armeriu maritima seeds (Plumbaginaceae) by Fowden (1958). Seed from Peganurn h u n m l a (Zygophyllaceae) contains this amino acid in large amounts (Table V). Based o n tentative identification, the compound occurs also in seed from Culliandrrr eriopl7j.lla and Plantago ocata of the families Leguminosae and Plantaginaceae, respectively (VanEtten et a/., 1963a). Isolation or tentative identification from four plant families indicates the componnd t o be more widespread than other less familiar amino acids, such as those only in Leguminosae. Examples are canavanine (see above), albizziin, 2,3diaminopropionic acid, mimosine, and willardiine in the subfamily Mimosoideae (Gmelin, 1959) and free amino acids and related compounds implicated in lathyrism (Schilling and Strong, 1955; Ressler er ai., 1961 ; Ressler, 1962). Some species from the Leguminosae. plants which commonly live symbiotically with nitrogen-fixing microorganisms, contain these less familiar amino acids and related compounds, all high in nitrogen.

NONPROTEIN XITROGEK

Table IV.

Sources of Nitrogen Other than from Amino Acids Found in Protein. In many seed meal hydrolyzates a relatively low percentage of the nitrogen is in the amino acids of protein (see Table 111, next to last column). The ammonia nitrogen probably is formed in great part from the amides of glutamic and aspartic acids. Other sources of ammonia are threonine and serine, known to be unstable to acid hydrolysis. Such unstable amino acids include less familiar amino acids from the seed of some plant species, 1086 J. AGR. FOOD CHEM.

Canavanine Content of Seeds from Leguminosae Canavanine, Grams per 16 Genus and Species Grams h-itrogen Ahrits precutoriirs 0.7 Astragalus pcnidirrutirs 12.9 Hedysariirn ,forzttrtresii 9.6 Melilotiis alba 0.7 Srorpiirrus sitbcillosri 7 .I Viciri angiistifolitr 4.3 Vicia giganterr 11.3

Table V.

Major Elution Peaks from Unidentified Compounds and Less Familiar Amino Acids

Genus and Species Ahrirs precatoriiis

Family Leguminosae

Akebiii trifoliata

Lardizabalaceae

Aspitodeliis microcarpiis Astragaliis patidura f us Baptisia leiicaritlia Cohaea scarideris Dalea r!iitaris Dolichos Iablah Eriogotiirt~nlatririi

Liliaceae Leguminosae Leguminosae Polemoniaceae Leguminosae Leguminosae Polygonaceae

Hrdysariirtn fori fariesii Lrippiilti redo wsk ii

Leguminosae Boragiiiaceae Leguminosae

Lntliyrirs s~~lrertris

Cucurbitaceae Compositae Cucurbitaceae Loasaceae Zygoph yllaceae Leguminosae Leguminosae

Elution

Positiona 0 . 73c 1.03 1.10 0.99 0.71 1.14 1.07 1.09 1.03 1.07 1.14 1.07 1.14 0.74c 1.11 1.13 1.12 1.03 0.72c 1.13 1.03 0.98 1.03 1 .05c 1.11 1.11 1.08 1.lld 0.94 1.04

Amount? Grams per 16 Grams Nitrogen 0 6. 0 1 0 2d

01 1 5 0 2

01 0 2d 0 0 0 0

3 3 1 1

0 4 3 8? 6 9 0 1

0 0 0 0 0 0 0 26 0

1 1 9e 1

2 7

2 8g

2d

0 2

6 0 0 5 01

Leguminosae Trigoriellri arrrhira Virrcci rose(i 1 1 Apocynaceae YR,,,,,,,,, $ , i l l = nil. of effluent of unknown peak divided by ml. of effluent of amino acid. Calculated as leucine, if absorption maxima was a t 570 mp, and as proline, if absorption maxima was at 440 my except for conditions dc-

scribed under (see belo\\). Absorption maxima at 440 mp instead of 570 m p . Possible identity 2,4-diaminobutyric acid. e Possible identity, 3-hydroxyproline. f Identified a s 2,4-diaminobutyric acid (VanEtten and Miller, 1963). Identified as 4-hydroxy-L-pipecolic acid based on identical elution position as the authentic compound on ion-exchange column chromatography and paper chromatography with I-butano1:ethanol:water 4 : 1 :4 upper phase and also based on nitrogen content, which agreed a i t h the theory for the compound 011 a small sample isolated in crystalline form. Amount present calculated from absorption color constant at 440 m p for the pure compound.

The unidentified peak R,,,,,. 0.72 to 0.74 with a 440mp maximum found in three species (Table V) could be due t o 3-hydroxyproline, recently isolated from acid hydrolyzates of sponge and collagen and shown to elute in this region upon ion exchange chromatography by Irreverre et td. (1962). Since several sugars and related compounds containing n o nitrogen pass through the column ahead of aspartic acid (Zacharius and Talley, 1962), further work is required to identify the peak. POTENTIAL O F SEED PROTEIN FOR F O O D O R FEED

Quantity of Protein. Large amounts of protein either before or after oil extraction are in seed meals from nearly all the 379 species---e.g., Table 11, columns 2 and 3. Of the species in Table I1 over 100 contain 2 5 % or more seed oil. These may be good sources of both protein and oil. For most species the amino acids in protein and ammonia, a high percentage of which likely originates from amide nitrogen (last two columns in Table II), are the source of more than 8 5 % of the crude protein nitrogen. This basis is a more meaningful measurement of protein

content, because nitrogen from other sources, if present, is not taken as a part of the crude protein. Seed from most of the species could well serve as a good source of low-cost protein if the plants could be as efficiently grown, harvested, and marketed as are our present major crops. Because of their high protein and lysine content, many are potential sources of protein concentrates to supplement cereals and starchy tubers. now major food sources especially in developing nations of the world. Protein Quality Based on Nutritionally Essential Amino Acid Content. In Figure 1, protein quality of seed meals is gaged by comparison of their nutritionally essential amino acids with recommendations for man by F A 0 (1957). For comparison the amounts of essential amino acids in corn, rice, wheat, and soybeans are also shown. Of seed meals from the 379 species, more than half are nutritionally adequate in lysine. Lysine deficiency is apparently a practical problem because of the relatively small amount of this amino acid in cereals, a major source of food and feed protein. Since cystine was not accurately determined for most of the 379 species, values for this amino acid are not included in Figure 1. The requireVOL. 15, NO. 6, N0V.-DEC.

1967

1087

that these oil seeds should provide a good supplemental protein t o use with cereal grains. The lysine content of these two Cruciferae is close to that for soybeans and the sulfur amino acids are present in larger amounts. The isoleucine content is lower. The pattern for crambe protein is much the same as that of hen’s egg except that crambe meal contains less isoleucine, As in Figure 1, the high leucine content of corn is apparent, a factor which, together with the low lysine, tryptophan, isoleucine, and valine, shows the unbalanced pattern of the essential amino acids in this grain. The opaque-2 corn under development is much improved, especially in lysine content. From information given in Table VI and Figure 1 and from examination of the amino acid composition of seed from individual species (Table I1 and from earlier publications), it is apparent that nearly all have potential as a protein source for food or feed. From the 379 species examined, and from the amino acid composition of seed from 54 species of the Gramineae reported by Taira (1962a,b, 1963), it appears that in his selection of the cereal grains, man picked seed proteins of poor nutritional quality. However, because of his omnivorous nature, the cereal grains probably serve as a good supplement to animal protein. Harmful Substances in Plant Seeds. .A major problem in the use of plants as a source of food or feed is t o remove or inactivate acute or cumulative toxic substances if present. Those substances that exert a delayed or cumulative effect are a greater hazard, because such effects are often hard to recognize and to relate to the cause. In a search for new plant seed as a food or feed, a major effort should be placed o n detecting and removing any possible toxins. The nonprotein nitrogenous substances in the seed of many Legiiminosae are known or suspected to be toxic. The thioglucosides present in the Cruciferae and related plants are the source of goitrogens (Greer, 1962). Selection of

ment shown for methionine is based on the assumption that enough cystine is also present to replace a maximum amount of the methionine. Even with this optimistic method of evaluation, the mean for methionine for the 379 species and for the cereals and soybeans are all below the minimum requirement. By calculations from the lower sulfur amino acid requirement recommended by F A 0 (1965), rice and corn are adequate in methionine. Isoleucine is also deficient in seeds from most of the 379 species. Soybean meal is high in this amino acid, as well as in lysine. Of the remaining amino acids, except for tyrosine, most seeds have more than adequate amounts. However, wheat is lower in threonine than the recommended requirement. In comparison with the provisional amino acid pattern, cottonseed is low in methionine and isoleucine (Carter er d., 1966). The nutritionally essential amino acid requirements for swine (NAS-NRC, 1959) are similar to those of man. Poultry requirements are higher and include arginine, glycine, and histidine as essential amino acids for optimum growth (NAS-NRC, 1960). Protein Quality of Selected Seeds Compared with the 1965 FAO/WHO Pattern. The F A 0 provisional amino acid requirement for man was modified (WHO, 1965). More emphasis is now placed on the pattern of the essential amino acids. The requirement for each essential amino acid is expressed in milligrams of the amino acid per gram of the total essential amino acids in the protein source. Such a calculated pattern of the essential amino acids in whole hen’s egg is used as a standard because its pattern satisfies the b-st estimates. Essential amino acid patterns for a number of major seed protein sources and for Crctnibe ubjsssinica and Lesquerellu seed selected from the 379 species are given in Table VI. The amino acid pattern and the high-protein content of defatted seed from C. ubyssinica and Lesquerellu indicate ~

Amino Acid Lysine Total sulfur-containing amino acids Methionine Cystine Isoleucine Leucine Total aromatic amino acids Phenylalanine Tyrosine Threonine Tryptophan Valine E/T ratio,i g./g. Crude protein,

Hen’s Egg

125

Table VI. Essential Amino Acid Patternsa A/E ratio. Milligrams per gram of total essential amino acids LowOpaque-2$ Protein Corn” Corn Ricec Soybeansri \\‘heate Crambef 66 117 111 157 80 140

Ixsquerellag 180

107 61 46 129 172

75 47 28 94 328

82 47 35 95 244

76 53 23 114 223

74 37 37 118 171

126 47 79 106 205

126 47 79 106 170

99 38 61 99 159

46 52 1 I3 177

195 114

217 123 94 85 17

209 120 90 97

186 109 76

137 2.51 10.6

205 115 90 99 30 126 2.71 61 .O

233 142 91 87 32

1 I8

205 144 61 96 33 142 2.47 7.3

178 99 79 115 35 135 2.16 31 . C

22 1 132 89 98 27 162 2.04 60.0

81

99 31 141 3.22

...

2.65 10.5

...

130

2. I O 15.9

113

33 128 2.25 49.0

Calculated as described by W H O (1965). From data of Mertz et al. (1965) except for tryptophan. Tryptophan content taken from Block and Wciss ( 1956). From data of Cagampang et al. (1966). High-protein rice gave a similar pattern, except lysine was 87 and tryosine, 85. From data of Rackis et al. (1961). e From data of Waggle et al. (1967) consisting of average from nine varieties of spring and \\.inter wheats. From data of Miller e f a[. (1962a) and VanEtten et al. (1961b). u From mean amino acid content from seed of 14 species of Lesquerella (Miller et al., 1962b). ’; From data of VanEtte? et a!. (1963b). Grams of essential amino acids per gram of nitrogen in seed meal. J’

1088 J. AGR. FOOD CHEM.

Safflower’, 98 98

seed from these plant families for food or feed uses would require extraction or other ways of removing these deleterious substances if they were present in large enough amounts to be harmful. In practice, seed meals may frequently be improved for feed or food uses by inactivation of deleterious substances by heat, mechanical separations, and selective extraction. Three examples can be cited of applications of these procedur-.s. Inactivation of deleterious enzymes and related substances by heat has been effective in toasting soybeans to inactivate trypsin inhibitors and hemagglutinins. This method of improving the product requires careful control, since excessive heat decreases the nutritional quality of protein for monogastric animals. The available lysine- and sulfur-containing amino acids are decreased by excessive heat treatment (Liener, 1958). Often a protein concentrate can be obtained by removal of indigestible hull. Safflower, as harvested, contains 45 % hull that makes it unsuited for monogastric animals. Mechanical removal of the hull and extraction of the oil give a meal containing 6 0 x crude protein of which more than 90% of the nitrogen is present as amino acids and amide nitrogen (VdnEtten et a/., 1963a). The removal of thioglucosides intact or as their hydrolysis products by solvent extraction gives a nontoxic meal from Crumbe uhj.ssinica seed (Tookey et a/., 1965; VanEtten et nl., 1966). Rats fed 28% of this meal in the ration for 90 days grew at a n essentially normal rate with n o evidence of toxicity based on histological examination of body organs (Tookey et ul., 1965). The extracted meal was higher in crude protein than the original defatted meal before solvent removal of thioglucoside products. Without extraction, about 87z of the crude protein nitrogen was derived from protein amino acids; after extraction. 97 % of the crude protein was derived from protein amino acids. The amino acid pattern was essentially the same before and after extraction, as the amount of each amino acid in the crude protein was increased by the extraction due to the removal of nonprotein nitrogen. ACKNOWLEDGMENT

The authors thank C. E. McGrew and B. R. Heaton for micro-Kjeldahl determinations; R. Gmelin of Biochemical Institute, Helsinki, Finland, for a sample of isolated 4hydroxypipecolic acid; R. W. Miller for some of the amino acid analyses; I. Stein for many of the computations; and I . A. Wolff for encouragement and advice. LITERATURE CITED

Altschul, A. M.. “Proteins, Their Chemistry and Politics,” Basic Books, New York. 1965. Anal. Chem. 33, 480 (1961). Bell, E. A,. Tirimanna, A. S. S.. Biochein. J . 97, 104 (1965). Benn, M. H., Cni7. J . Cliem. 43, 1 (1965). Birdsong. B. A,, Alston. R., Turner, B. L.. Cali. J . Botany 38, 500 ( I 960).

Block, R. J.. Weiss, K. W., “Amino Acid Handbook.” 1st rd.. Charles C. Thomas, Springfield, Ill., 1956. Cagampang. G. B., Cruz, L. J., Espiritu, S. G., Santiago. R. G., Juliano, B. O., Cereal Chem. 43, 145 (1966). Carter, F. L., Castillo, A. E., Frampton, V. L.. Kerr, T.. Phytochein. 5, 1103 (1966). Chisholm, M. D., Wetter, L. R., Curl. J . Biocliein. 42, 1033 (1964). Deyoe, C. W., Shellenberger,J. A., J. AGR.FOODCHEM.13, 116 ( I 965). Earle, F. R., Jones. Q., Ecori. B o t a ~ y16, 221 (1962). Food Agr. Orguri. U.N., F A 0 Niiir. Studies 16 (1957). Food Agr. Organ. U. N . “Protein at the Heart of the World Food Problem,” Rome, 1964. Fowden. L., Biochem. J . 70, 629 (1958). Gmelin, R., Hoppe-Seylers 2. Pliysiol. Cl7em. 316, 164 (1959). Greer, M. A,, Receiit Progr. Hormoiie Res. 18, 187 (1962). Hamilty, H., Ed., “World Population and Food Supplies, 1980, American Society of Agronomy, Madison. Wis. A S A Spec. Pub/. 6 (1965). Irreverre, F., Morita, K.. Robertson, A. V., Witkop. B. Biocliein. Biophys. Res. Commwi. 8, 453 (1962). Jones, Q., Earle, F. R., Ecoii. Boratiy 20, 127 ( I 966). Kjaer, A , , “Modern Methods of Plant Analysis.” L. Zechiiieistrr. Ed., p. 122, Springer-Verlag, Vienna, 1960. Liener, I. E., “Processed Plant Protein Foodstufs.“ A. M. Altschul, Ed., p. 79, Academic Press. New York. 1958. Meister. A,, “Biochemistry of the Amino Acids,’’ p. 2. Academic Press, New York, 1965. Mertz, E. T., Vernon. 0. A,. Bates, L. S.. Nelson. 0. E.. Scierice 148. 1741 (19651. Miller. R. W., VanEtten. C. H., McGrew. C.. Wolff. I. A,. Jones. Q., J. AGR.FOODCHEW10, 426 (1962a). Miller, R . W.. VanEtten. C. H.. W o l f , I . A,,J . A m . Oil Chemists‘ SOC.39, 115 (1962b). Natl. Acad. Sci.-Natl. Res. Counci!, Washington. D.C., “Nutrient Requirements for Poultry, Pub/. 827, 7 ( 1960). Natl. Acad. Sci.-Natl. Res. Council. “Nutrient Reauireinents for Swine.” Pirbl. 648, 12 ( I 959). Rackis. J. J., Anderson, R . L.. Sasanie, H. A,. Sniith. A. K.. VanEtten, C. H.. J. AGR.FOODCHEW9, 409 (1961). Ressler. C., J . Biol. Clieni. 239, 733 (1962). Ressler. C., Redstone. P. A,. Erenberg. R. H., Scie/icr 134, 188 ( I 961). Schilling, E. D., Strong. F. M., J . Am. Chem. Soc. 77, 2S13 (1955). Spackman, D. H.. Stein, W. H.. Moore, S., AIINI.Clir/n.30, I 190 f 1958). Taka, H,,Botuii. Mug. (Toohyo)75, 80 (1962a). Taira, H., Boruii. Mag., (Tokyo)75, 242 (1962b). Taira, H.. Botuii. Mug. (Tokyo) 76, 340 (1963). Tookey, H. L.. VanEtten. C. H.. Peters, J. E.. Wolf. I . A,. Cereal Cliern. 42, 507 (1965). VanEtten, C. H.. Daxeiibichler. M . E., Peters, J. E., WoltY. I. A,. Booth, A. N.. J. AGR.FOODCHEM. 13, 24 (1966). VanEtten, C . H., Miller, R. W.. Econ. Botaiiy 17, 107 (1963). VanEtten, C. H., Miller, R. W.. Earle. F. R.. Wolff. I . A.. Jones, Q.. J. AGR. FOODCHEM. 9,433 (1961a). VanEtten. C. H., Miller, R. W.. Wolff. I. A , . Jones. Q.. J . A G K . FOODCHEM. 9,79 (1961b). VanEtten, C. H., Miller, R. W., Wolff. I. A. Jones. Q.. J . AGR. FOODCHEW.. 11, 399 (1963a). VanEtten, C. H., Rackis, J. J.. Miller, R. W., Siiiith. A . K., J. AGR. FOODCHEM.11, 137 (1963b). Virtanen. A. I., Gmelin, R., Acta Cliem. S c u d 13, I211 (1959). Waggle, D. H., Lambert. M. A,, Miller, G. D.. Farrel. E. P.. Deyoe, C. W., Cered Cliem. 44, 48 (1967). World Health Organization. “Protein Requirements.” Report of a Joint FAOIWHO Expert Group, Rome, W H O Tech. Rept. Ser. 301 (1965). Zacharius, R. M., Talley, E. A.. J . Cliromatog. 7, 51 (1962). Receiced .for reciew Jirne 23. 1967. Accepted J d y 16. 1967. Dicisioii of Agricirltiirul uiid Food Chemistry, 153rd .Weering. ACS, Miuini, Floridu. April 1967. Meiitiori of ,firm ri(inics or /rude products does 1701 rmistitirte uti endorsemenr b y [lie USDA orer other,firnis or .\iinilur prodircrs iiot nietitioned.

VOL. 15, NO. 6, N0V.-DEC. 1967

1089